It's taken me five years of on-again, off-again but often substantial playing of poker (don't worry, mostly not cash games) to really understand some of the game's concepts that I read about when I was playing but didn't properly understand.
Ludological scholars are right to insist, in this respect and others, that games require attention as games, that they have a character or nature that is intrinsic to games and not to texts or performances or sociality. Poker has a "deep game" that is not spelled out in the rules, but which powerfully sorts out the losers and the winners, given a sufficient number of rounds of play.
What's the "deep game" of virtual worlds, those that have at least some game-like character?
Poker really has two deep games that are not visible in its explicit rules. The first doesn't take a lot of play to understand or appreciate, and some players can never get good at it because they lack the capacity to do so. That's the ability to read another player's patterns of play and emotional posture, to look for his or her characteristic "tells". This is a cultural game, a social game: if you're good at it, it's because you can import sociality or emotional intelligence into the infrastructure provided by poker's rules. You can even do it online, where there are no faces or bodies to look at. The game's rules still play a role in using this skill correctly, and so does luck in terms of where the loose and tight players in a game are in relationship to oneself. I have gotten a bit better over time in reading other players, but people can be geniuses at that aspect of the game in a day, or never get any better at it in a decade.
The other deep game has to do with the value of position. This is structural, it's a consequence of the ruleset, but it's not visible in the rules per se. Every poker guide and handbook you can read will lay out this aspect of the game, but I think it is very hard to understand fully how and why it matters until you have played a great deal, and played in games where players have at least some respect for the stakes on the board. (E.g., the chips are real money, or played for real money.) In hold'em poker, there are hands you simply don't want to play ever from some positions that might be worth playing aggressively from another position, in dynamic relationship to the size of your own stack of chips versus others on the board.
Virtual worlds have a sociality game deeply embedded within in them, obviously. And as with poker, some players excel at this game from the beginning, in a variety of ways. Both scammers and guild leaders may be excellent social players, in their own fashion. Some people will never get good at it. But precisely because virtual worlds are so robust in their social dimensions, I think it's right to argue, as Constance Steinkuehler, T.L. Taylor and many other scholars have argued, that virtual worlds actually teach sociality, that many players improve in some dimensions of their social intelligence over time: in their functioning within organizations, in their coordinated response to collective action problems, in setting personal goals and achieving them through social networks, in communicating with other individuals. We can make fun of the dysfunctionality of a lot of people playing in these worlds, and even thirty minutes spent in the Barrens tends to make one feel that some players are losing rather than gaining social intelligence in virtual worlds. But I still think the social context of even a simple virtual world runs along so many more axes than poker, and is so much more mimetic to the real world, that this deep game can be learned quite well and often is. (The deep social game of poker, in the end, is limited almost entirely to performing lies and reading lies. Unless you're cheating or flirting, it's doesn't involve other kinds of emotional or social connection to the players. )
The other deep game, however, strikes me as rather like poker's: it is about the hidden, emergent, or unintended consequences of the rules or code that govern play in the world. This is where most of the discourse about "cheating" in virtual worlds lies, and where most of the moral debate about how they are meant to be played ultimately centers. Players frequently discover that some aspect of play has unintended consequences that are disproportionately advantageous. A character's powers can be used in some novel fashion that renders that character nearly invincible versus certain opponents. A computer-controlled opponent has some unexpected vulnerability that makes it a risk-free source of reward. A tactic interacts with the virtual physicality of the landscape in some surprising fashion.
Sometimes this behavior is straightforwardly banned or forbidden, and the game's code amended to prevent it. Sometimes the designers compliment the players for having discovered this deep aspect of play, and it rapidly becomes the new standard. Not too long ago, my own World of Warcraft guild was in Karazhan, working on some new trash mobs as we climbed up the tower, and somewhat coincidentally, we found out that a particular tactic that was not mentioned in WoWWiki removed some of the need for careful coordination around clearing them. (The tactic is now in WoWWiki, I noticed.)
So does this "deep game" require a lifetime to master? Yes and no. It does in the sense that you have to have played two or more of the commercial virtual worlds that have come out since Meridan 59 to understand where the opportunities for this kind of play are likely to exist, and more importantly, how to protect knowledge about these kinds of discoveries for as long as possible while also being knowledgeable about what kinds of discoveries may trigger designer intervention and even designer punishment, relative to the established behavior of a given developer. (E.g., in the first Asheron's Call, you could pretty much try and get away with anything without fear that the developers would punish you for it, whereas Blizzard is known to crack down fairly hard on players discovering unforeseen or emergent aspects of the gameworld in many cases.)
The problem is that because virtual worlds are almost entirely built on the same basic rule-structure derived from DikuMUD, and because their representations of physical and graphical environments are ultimately so similar, this deep game becomes more and more known to larger and larger numbers of players over time, all the more so since World of Warcraft has evolved into the new template for all subsequence virtual-world games. But unlike poker, this is not a deep game that opens up a vast new domain of contingent decisions for players that help to further distinguish or individuate them in their style of play. Once you fully understand how position works in poker, you do not play the same as everyone else who understands position. You simply make better decisions about your own style of play, craftier decisions about how to take the kinds of risks you want to take, and how to apply the deep game of position to the deep game of reading the psychology of other players.
The deep game of "find the unintended aspects of play and arbitrage the living hell out of them" in Diku-style worlds, on the other hand? It actually forces a convergence of play styles and limits the social variety of players over time, which is why the debate about this form of play stops being about rules and starts being about morality. That alone is a good reason for designers to think about the Diku template again. When players discover the deep game, you want it to be a new source of fertility, not a barren monoculture.